%0 Conference Proceedings %T Kronecker Decomposition for GPT Compression %A Edalati, Ali %A Tahaei, Marzieh %A Rashid, Ahmad %A Nia, Vahid %A Clark, James %A Rezagholizadeh, Mehdi %Y Muresan, Smaranda %Y Nakov, Preslav %Y Villavicencio, Aline %S Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) %D 2022 %8 May %I Association for Computational Linguistics %C Dublin, Ireland %F edalati-etal-2022-kronecker %X GPT is an auto-regressive Transformer-based pre-trained language model which has attracted a lot of attention in the natural language processing (NLP) domain. The success of GPT is mostly attributed to its pre-training on huge amount of data and its large number of parameters. Despite the superior performance of GPT, this overparameterized nature of GPT can be very prohibitive for deploying this model on devices with limited computational power or memory. This problem can be mitigated using model compression techniques; however, compressing GPT models has not been investigated much in the literature. In this work, we use Kronecker decomposition to compress the linear mappings of the GPT-2 model. Our Kronecker GPT-2 model (KnGPT2) is initialized based on the Kronecker decomposed version of the GPT-2 model and then is undergone a very light pre- training on only a small portion of the training data with intermediate layer knowledge distillation (ILKD). Finally, our KnGPT2 is fine-tuned on downstream tasks using ILKD as well. We evaluate our model on both language modeling and General Language Understanding Evaluation benchmark tasks and show that with more efficient pre-training and similar number of parameters, our KnGPT2 outperforms the existing DistilGPT2 model significantly. %R 10.18653/v1/2022.acl-short.24 %U https://aclanthology.org/2022.acl-short.24 %U https://doi.org/10.18653/v1/2022.acl-short.24 %P 219-226